AI functions in the optimization of design
| AI functions (transcription codes) | Selected quotes |
|---|---|
| Optimizing building designs | “Use of computational design by training a machine learning (ML) model to optimize or predict design elements. Taking data from the previous designs to train the model and design algorithms.” (Int #7) “AI allows designers to significantly optimize the design.” (Int #1) “Use AI as a statistical tool to help design not necessarily on the visual aspect, but maybe some optimization that you can do with AI.” (Int #4) “Deploy computational design methodologies and use optimization algorithms to come up with design solutions.” (Int #6) “If you take like the recipes from the previous designs and you train a computer model based on all of that prior experience, the model is able to predict or produce an optimized design.” (Int #7) “It can give you the pros and cons, and it could suggest spatial rearrangements of buildings for optimization.” (Int #8) “We created kind of fitness functions so we could tweak what the AI or the algorithm considered good or bad designs. It would be able to analyze and assess them.” (Int #6) |
| Suggesting energy-efficient designs | “Computational design can suggest energy efficient design.” (Int #3) |
| Generating building design | “Use AI in apartment layouts and the floor plans and giving you lots of options to pick from as a designer.” (Int #2) “Image generators for inspiration, maybe Midjourney is one that I find more inspiring imagery, and it’s a bit more difficult to use. [Moreover] Grasshopper or Dynamo for automation design.” (Int #3) “There are some tools out there like Layout. Having the whole host of things that are doing like internal design layouts and usually using AI.” (Int #8) “Using genetic generative and adversarial network to create new data for assisting in generating building layouts and floor plans. It is all done by Midjourney.” (Int #6) “High pub is an online computational platform, which has created a prompt-based, generative building designer.” (Int #3) “If you want to design a product, we can use AI or ML to have different design options. And based on the LLM, we can produce some factors.” (Int #5) “Generative design is already well established in CAD tools like Autodesk. This is not a data challenge; it is more about setting up vectors or ranges of constraints in graphs and for things like human interaction, weather, earthquakes, and physics for load-bearing or longevity.” (Int #9) “I think that we can create different graphs, different progress, or different design solutions based on how you modify certain constraints.” (Int #4) “The current computational design is more about the computer being used to generate a generative AI process to generate permutations and combinations of designs, then assess them and determine which is the best one based on previous data.” (Int #7) |
| Optimizing building cost | “Use AI to look at what the available materials are and try and run generative models to see how much the cost of potential developments could be impacted”. (Int #3) |
| Optimizing structural engineering calculation | “Make the structural engineering calculation process much more efficient through some sort of ML model.” (Int #2) |
| Simplifying/improving modeling processes | “Use of AI simplifies computationally complex modeling processes.” (Int #1) “Within the software (use ML techniques), we streamline a whole range of really demanding computationally demanding routines, thus improving our modeling capabilities.” (Int #1) |
| Simulating loading conditions | “Use AI to simulate the loading condition of the specific design based on the lifetime of our current data and information.” (Int #2) |
| AI functions (transcription codes) | Selected quotes |
|---|---|
| Optimizing building designs | |
| Suggesting energy-efficient designs | |
| Generating building design | |
| Optimizing building cost | |
| Optimizing structural engineering calculation | |
| Simplifying/improving modeling processes | |
| Simulating loading conditions |
Source(s): Authors’ own work
Sharing content requires targeting cookies to be enabled. Please update your cookie preferences to use this feature.